ML19344D515
| ML19344D515 | |
| Person / Time | |
|---|---|
| Site: | Yankee Rowe, Maine Yankee |
| Issue date: | 02/29/1980 |
| From: | Vandenburgh D YANKEE ATOMIC ELECTRIC CO. |
| To: | Denise R Office of Nuclear Reactor Regulation |
| References | |
| RTR-NUREG-0630, RTR-NUREG-630 WMY-80-33, WYR-80-27, NUDOCS 8003120575 | |
| Download: ML19344D515 (4) | |
Text
r 0
Telephone 6f7 366-91 11 rwx 710-390-0739 YANKEE ATOMIC ELECTRIC COMPANY s.3. u WYR 80-27 WMY 80-33 V
20 Turnpike Road Westborough, Massachusetts 01581
_lar
.YAmxes
~~-
February 29, 1980 United States Nuclear Regulatory Commission Washington, D. C.
20555 Attention: Office of Nuclear Reactor Regulation Mr. Richard P. Denise Acting Assistant Director for Reactor Safety
Reference:
(1) License No. DPR-3 (Docket No. 50-29).
(2) License No. DPR-36 (Docket No. 50-309).
(3) USNRC (R. P. Denise) letter to YAEC, Re:
" Request for Comments on Temperature Ramp Rate Calculations for Cladding Swelling and Rupture Models", January 31, 1980.
(4) A. Husain et al., " Application of Yankee - WREM - Based Generic PWR ECCS Evaluation Model to Maine Yankee", YAEC-1160, July, 1978.
(5) USRRC letter to YAEC, Re:
" Evaluation of Topical Report YAEC-1160", January 17, 1979 (6) D. A. Powers and R. O. Meyer, " Cladding Swelling and Rupture Models for LOCA Analysis", Draft NUREG 0630, November 8,1979 (7) D. A. Powers and R. O. Meyer, " Cladding Swelling and Rupture Models for LOCA Analysis", Draft NUREG 0630, February 12, 1980.
(8) XN-76-27, " Exxon Nuclear Company WREM-Based Generic PWR ECCS Evaluation Model Update ENC-WREM-II", July, 1976.
(9) YAEC letter to USNRC (D. G. Eisenhut), Re: " Evaluation and Verification of Cladding Swelling and Rupture Models for Yankee Rowe LOCA Analysis", January 7,1980.
(10) MYAPCO letter to USNRC (D. G. Eisenhut), Re:
" Evaluation and Verification of Cladding Swelling and Rupture Models for Maine Yankee LOCA Analysis", January 7, 1980.
(11) YAEC letter to USNRC (R. P. Denise), Re:
" Technical Review of Draft NUREG 0630", December 10, 1979 (12) ORNL letter (R. H. Chapman) to USNRC (K. Eniel), Re:
" Comments on Draft NUREG 0630", January 14, 1980.
Dear Sir:
Subjec t: Comments on Temperature Ramp Rate Calculations for Cladding Swelling and Rupture Models This letter provides comments on the issues presented in your communication of January 31, 1980 pertaining to the inclusion of cladding heatup temperature ramp rate effects in hg strain and rupture anaAysis of Zircaloy cladding behavior during a LOCA 31 The issues addressed include a review of cladding heatup temperature ramp rate calculational methods in YAEC
~~
8003120
Y U. S. Nuclear Regulatory Commission February 29, 1980 Page 2 fuel performance codes, comments on the calculational methodology presently derived for the NRC audit code as described in the referenced communication, and comments on the calculation of pre-ruptured cladding plastic strain in LOCA analysis.
The models used by YAEC for cladding swelling and e in PWR LOCA analysis have been described in previous transmittals.({ugt e e To investigate the impact of incorporating cladding heatup temperature ramp rate effects in the analysis in response to({>C requests, the draft correlations in the November issue of Draft-NUREG-0630 were applied in break spectrum, depletion, and power shape parametric studies by employing the following general procedures:
(a) Cladding temperature versus time curves from the reload license calculations were examined for the adiabatic heatup/reflood portion of the transient to determine typical cladding heatup rates during plastic deformation prior to rupture. The nominal rates were N SOC /sec for 0
Yankee Rowe Cycle 14 and N 10 C/sec for Maine Yankee Cycle 5.
These rates remain essentially constant for the rupture node prior to rupture although variations exist in some cases with relatively long time-to-ruptu re.
For example, the Maine Yankee analysis showed some ramps decreasing to N 3 C/see as cooling and deformation progressed.
(b)
Correlations for rupture temperature versus rupture pressure w by interpolation of the curves in the November draft document (gre derived
) in each investigation using the nominal cladding heatup rates.
Subsequently, these correlations were found to compare satisfactorily to those obtained directly from the QRNL formulation presented in the more recent issue of Draft-NUREG-0630.(7)
(c) These correlations and those in Reference 6 for low cladding heatup ramp rates for cladding strain versus cladding temperature at rupture and for local flow blockage versus cladding rupture temperature were used in place of the licensed models in the YAEC {00DEE-2 heatup analysis.the Yankee Rowe For associated with 20% reduction in flow area was employed, consistent with the prediction of the low temperature ramp rate local blockage curve.
The results of these analyses have been reported in previous transmittals. t 9,10 )
(d) As a final check on these procedures, cladding temperature versus time curves for these analyses were compared to those obtained from the original reload licensing calculations. Typical cladding heatup rates during plastic deformation prior to rupture showed no significant change; therefore, further iterations of the procedure were deemed unnecessary.
YAEC continues to support this type of approach for the evaluation of impacts resultant from the incorporation of cladding heatup ramp data effects in LOCA fuel rod analysis. Whereas a mechanistic approach to cladding behavior analysis may suggest the type of modeling application of ramp rate outlined in the modifications to the NRC T00DEE2 audit code (i.e., biased
f U. S. Nuclear Regulatory Commission February 29, 1980 Page 3 averaging techniques applied to instantaneous cladding heatup rates over the regime of plastic deformation), it is not clear that this approach addresses the following concerns:
(1) Cladding heatup ramp rates reported in the data base are generally those experienced during the heatup ramp prior to the time-of-burst.
Ccnsequently, as the burst time is approached in the Zircaloy low l
temperature (<95000) regime, instantaneous heatup rates are consitentl{21guch less than those reported during the preliminary heatup
{
ramp. I'll e Therefore, the data base does not support the complexities of the computational model definitions.
If the computitional model focuees averaging or biasing the ramp rates to those nearest the rupture temperature, the derived rupture parameters will not reflect the data in an appropriate manner.
In the analysis of the temperature range example above, the computational scheme will result in a ramp rate which is artificially low in comparision to the dc ta base and will predict correspondingly larger rupture strains and local flow blockage.
(2) Weighting the average cladding heatup ramp rate in favor of ramp rates nearest the rupture temperature does not appear appropriate when cladding is in a high temperature regime. The cladding heatup rate may have been low prior to and during the onset of plastic deformation, but may increase significantly near the rupture temperature due to enhanced n oxidation rates. As indicated in the NRC evaluation of the data cladd{7
- base,
" slow-ramp rates produce very small strains at temperatures greater than about 95000 because the Zircaloy has time to oxidize and embrittle before significant ballooning can occur" and " fast-ramp rates produce the opposite effects". Therefore, even though the cladding has experienced a low ramp rate and should experience low strain and local blockage at rupture, the mathematical model will be biased by
" instantaneous" ramp rates near the rupture temperature and will predict artificially large cladding strain and local blockage.
In conclusion, the model should be derived to function within the limitations of the data base. Since the computer model accesses and applies directly the correlations derived from the data base, the averaging technique for cladding heatup rate should be devised to best match those upon which the data base and correlations are founded. Utilizing a typical cladding heatup ramp rate to derive static tables of rupture temperature versus pressure, cladding strain, and local flow blockage appears to best meet current limitations.
The issue of incorporating time-at-temperature dependence of preruptured cladding strain suggests the application of an enhanced phenomenological modeling approach for rupture bahavior which integrates cladding mechanical performance, oxidation embrittlement, and mechanistic rupture prediction models. The discrepancies in cladding temperature heatup ramp rate characterization in measurement approaches and modeling applications indicated
~ by the issues addressed in the above items (1) and (2) demonstrate the
- f.,
6 U. S. Nuclear Regulatory Commission February 29, 1980 Page 4 attention required to correctly and conservatively incorporate such a complex and interactive modeling refinement. Initial efforts should focus on design specification of necessary parameters and measurement techniques to establish the required qualified data base.
With respect to the above discussions, we continue to believe that the evaluations, analyses, and conclusions reported in References 9 and 10 for Yankee Rowe and Maine Yankee, respectively, remain valid. We wish to emphasize the conclusion, especially applicable to the Yankee Rowe analysis, that continuing acquisition and evaluation of new data is required, particularly in the low cladding heatup rate /high cladding temperature region.
During 1980, Yankee will be approaching the NRC with a rod heatup model to replace T00DEE-2 as part of our licensed ECCS codes. This model will address all available data and current technical performance and modelir.g issues.
We trust that this information is satisfactory; however, if you have any further questions, please contact us immediately.
Very truly yours, Ch* k D. E. Vandenburgh Senior Vice President Yankee Atomic Electric Company SPS/smh i
i